The Relationship of Total Atmospheric Ozone to the Sunspot Cycle

For the 27-year period 1933-1959 inclusive, highly significant negative correlation is found between the relative sunspot number and the worldwide average of total atmospheric ozone. Peak correlation with respect to the present-day 10-year sunspot cycle is found at a lag of 1•/• to 2 years of the sunspots relative to ozone. The peak correlation of total atmospheric ozone with the mean latitudinal distance from the solar equator of the total area of sunspots is almost identical in magnitude to that with sunspot number. Itowever, the peak correlation with sunspot latitude is found at a lag of less than 6 months of the ozone relative to sunspot latitude, in the opposite direction to that with sunspot number. This phase difference of the negative correlation suggests that atmospheric ozone is much more sensitive to sunspot latitude than to sunspot number. INTRODUCTORY P•EMARKS The author's longtime interest in atmospheric ozone as a possible link between solar activity and climate was increased by the implications of a recent paper by Kraus [1961]. He concludes that of the five possible factors of climatic control which he considers, ozone, volcanic dust, solar constant, solar corpuscular radiation, and CO.o, the probable effects of increased atmospheric ozone appear to fit best the general pattern of climatic change in high and low latitudes during recent decades. Subsequent discussion with Professor C. E. Palmer and with Dr. S. P. Venkiteshwaran at U.C.L.A. served to further this interest and to indicate the probable adequacy of present ozone data for a preliminary statistical check of any possible solar-ozone relationship. Apparently no previous effort has been made to check this possibility on a large scale. A remark by Dr. Hans Diitsch at Boulder in April 1961 directed the author's attention to a compilation of all available northern hemispheric ozone data currently under preparation by Dr. J. London for Air Force Cambridge Research Laboratories. Dr. London was most obliging in supplying a complete tabulation of available data, in the form of monthly means of total ozone, for all observing stations of the northern hemisphere from the first observations at Oxford in 1925, up to and including 1959. Corresponding data from the southern hemisphere for Wellington, New Zealand, were kindly supplied by M. A. F. Barnett, Director of the New Zealand Meteorological Service, and for Aspendale (Melbourne), Brisbane, and Macquarie Island by C. H. B. Priestley, Chief of the Division of Meteorological Physics of the Scientific and Industrial Research Organization of the Commonwealth of Australia. All the ozone values, from all three sources, are based on the absorption coefficients found by Vigroux, and are expressed in the standard unit, milli-atmocentimeters; hence they should be strictly comparable. ]PRocEDURAL TREATMENT OF Tt-IE DATA Nature o/the data and averaging procedures. All the data used in this study were supplied in the form of monthly averages of individual station measurements of total ozone. In every case the number of individual daily observations that went into the station monthly mean was listed along with the mean value as tabulated. The great majority of station monthly mean values represent from 20 to 31 individual daily values, a substantial number of them contain between 10 and 20 daily values, and a small scattering contain less than 10 daily values. However, in averaging monthly mean station values to obtain either an annual mean value for a single station or a calendar monthly average value for a number of stations, in no case were the individual station monthly mean values weighted by the number of daily observations. This procedure was followed because the sampling in-